Recently, Adrian Murray and Brent Baiotto, Fisent’s Founder & CEO and COO, respectively, told me a story about a healthcare insurance company they are working with that is utilizing generative AI to automate its claims processing.

Fisent’s Applied GenAI Process Automation solution, BizAI, has enabled the healthcare insurance company to massively speed up the processing of claims, allowing them to reduce costs, enhance efficiency, and free up a lot of time for their assessors, while simultaneously improving customer service and experience.

While all of that is impressive, it isn’t what I believed was the truly intriguing aspect of the story.

That came when Murray and Baiotto told me what the healthcare company decided to do when they came across claims with issues that required resolution before they could be processed.

Historically, when insurers encountered claims like these, they would contact clients via email to request the missing information. Typically, this would initiate a back-and-forth email exchange with the client in search of the missing information, which could sometimes take days or even weeks to resolve.

However, instead of trying to also automate that part of the process, the healthcare company took a step back and asked themselves what would be the best and quickest way to address these issues for both themselves and their clients.

What they realised was that a technological solution would not necessarily be quicker than if their assessors simply picked up the phone and called their clients directly when they encountered an issue with a claim. This was made easier by the fact that the initial claims processing automation would free up a lot of their claim assessors’ time. They also concluded that an additional benefit of adopting this approach was that it would be an excellent way to strengthen their relationships with clients and demonstrate their care for them.

Now, the reason I like this story is that it shows how the healthcare company didn’t take a technology-first or AI-first approach but rather took an experience-first approach to the problem and kept their clients at the center of their decision-making.

This is not usual.

Too often do I hear stories about occasions where companies try to harness the power of new technologies, such as generative AI and automation, only for the customer to get lost or overlooked in the process.

Take, for example, the case of Klarna, which, a couple of years ago, went all in on AI and sought to automate most if not all of its marketing and customer service jobs.

Along the way, they claim to have saved millions of dollars in marketing costs, that their chatbot was responsible for two-thirds of customer service chats and was completing the work of 700 employees and that the hiring freeze that they initiated as part of this process has helped them reduce their workforce by 22%.

However, earlier this month, Klarna’s Founder and CEO Sebastian Siemiatkowski admitted that their shift to fully automating customer service was not going as planned and that cost reduction had played too big of a role in the formation of their initial plans. Consequently, they adopted an approach that ultimately resulted in lower-quality outcomes.

That doesn’t mean that their automation efforts were a complete bust. Not at all. By all accounts, their chatbot works quite well, but can’t deal with more complex or nuanced queries.

As a result, Klarna is now looking to hire more freelance gig workers to tackle these more complex customer queries.

What’s clear about the case of Klarna is that when they were considering what to automate in their business, they adopted a technology and AI-first approach, and any real consideration of what was best for the customer was crowded out by their overriding desire to reduce costs.

This wasn’t the case for Fisent’s healthcare insurance client. They took an experience-first approach, and always kept what was best for their clients at the center of their decision-making.

The lesson for other brands thinking about AI and automation is to always adopt an experience-first approach. This will help ensure that the customer remains front and center when making technology decisions.

This post was originally published on Forbes.com.

Credit: Photo by Miryam León on Unsplash